FaithTrace uses the directional derivative of class logits along text-induced directions in feature space as an influence score to produce and evaluate more faithful zero-shot textual explanations for image classifiers.
Title resolution pending
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
cs.CV 2verdicts
UNVERDICTED 2representative citing papers
TEXTER generates zero-shot textual explanations for image classifiers by isolating decision-critical features from contributing neurons, mapping them into CLIP space, and using sparse autoencoders for improved interpretability in Transformers.
citing papers explorer
-
Zero-Shot Faithful Textual Explanations via Directional-Derivative Influence on Predictions
FaithTrace uses the directional derivative of class logits along text-induced directions in feature space as an influence score to produce and evaluate more faithful zero-shot textual explanations for image classifiers.
-
Zero-Shot Textual Explanations via Translating Decision-Critical Features
TEXTER generates zero-shot textual explanations for image classifiers by isolating decision-critical features from contributing neurons, mapping them into CLIP space, and using sparse autoencoders for improved interpretability in Transformers.